How to Read a Purchase Intent Score

You finished your study. The dashboard shows a purchase intent score of 3.7. You're supposed to walk into a meeting tomorrow and say something intelligent about it.

Is 3.7 good? It depends. But probably not on what you think.

Start with the distribution, not the average

Every purchase intent score you see in Gustoso is the mean of a probability distribution across five points — from strongly negative (1) to strongly positive (5). That mean is a summary. It's the first number you see because it's the easiest to compare across studies, but it's the last thing you should interpret.

The distribution is the story. The average is the headline.

A 3.7 can come from a cluster of responses sitting politely around the middle. It can also come from a U-shaped split where half the panel loves it and half the panel rejects it outright. Same mean. Completely different product decision.

Before you look at the number, look at the shape.

Top-2-box and bottom-2-box

The industry shorthand for reading a Likert distribution is the top-2-box and the bottom-2-box — the combined share of responses at 4 and 5 (people leaning toward buying) and the combined share at 1 and 2 (people leaning away).

Top-2-box tells you your reachable audience. Bottom-2-box tells you who's actively turned off. The middle (3) is ambivalent — often the largest group, and often the least informative.

A 60% top-2-box with a 10% bottom-2-box is a strong product. A 45% top-2-box with a 35% bottom-2-box — even if the average is identical — is a polarizing one. Polarizing products can still win, but they're a different kind of bet.

What counts as "good"

There's no universal benchmark. A 4.2 on a snack product is modest. A 3.5 on a new financial service could be outstanding. Category norms vary too widely to name a single target.

What matters is relative comparison. Benchmark against your own work. Run two or three variants of the same concept. Whichever scores highest is your winner, and the gap between scores tells you how decisive the winner is.

If variant A scores 3.8 and variant B scores 3.7, you don't have a winner — you have a tie and some noise. If A scores 3.8 and B scores 3.2, A is meaningfully ahead.

Absolute scores sell decisions. Relative scores make them.

Distribution shape tells you more than the mean

Three shapes show up repeatedly, and each means something different.

Tight and centered — most responses cluster around a single point. This is a consensus product. Whatever the mean is, the panel mostly agrees with it. Safe, predictable, probably not a breakout.

Skewed positive or negative — a long tail on one side. This is a clear signal. A positive skew on purchase intent means you have genuine enthusiasts; a negative skew means you have genuine critics. Either way, the direction is real.

Bimodal — two peaks, often with a hollow middle. Polarizing. Certain segments love it, others reject it. Your next move isn't to tweak the creative — it's to figure out who the two groups are and whether the lovers are the audience you actually wanted.

The mean hides all of this. The distribution shows it immediately.

Reading the dashboard in the right order

When you open a Gustoso report, read it in this order and you'll extract more signal in less time.

1. The distribution. Shape first. Tight, skewed, or split?

2. Top-2-box and bottom-2-box. Positive reach and negative friction.

3. The mean. Use it to compare against other variants or other studies — not as a standalone grade.

4. Segment breakdowns. Where are your strongest and weakest audiences? This is where distribution shape becomes actionable.

5. Free-text responses. The scores came from natural-language answers. Read a handful. The why behind the number is often more valuable than the number itself.

What to watch out for

Don't chase tiny deltas. A 0.1 difference between variants is usually noise. Look for gaps that survive across multiple dimensions.

Don't over-interpret a single score. One study, one audience, one moment in time. Treat it as a data point, not a verdict.

Don't skip the free text. Every score on Gustoso is derived from a natural-language response. If you want to understand why a concept scored the way it did, the answer is sitting right there in the transcripts.

The short version

A purchase intent score is not a grade. It's a summary of a distribution, and the distribution is where the real information lives. Read the shape, read the top- and bottom-2-box, read the text — and only then read the average.

Start a study and try reading one of your own reports this way. You'll notice things you missed before.

Photo by Stephan Halbach on Unsplash